a b s t r a c t

Purpose: Understanding differences between patients who accept and decline HIV testing is important for developing methods to reduce decliner rates among patients at risk for undiagnosed HIV. The objectives of this study were to determine the rates of acceptance and reasons for declining, and to determine if differences exist in patient or visit characteristics between those who accept and decline testing.

Basic procedures: This was a retrospective medical record review of all patients offered an emergency department (ED) HIV test from 11/1/11 to 10/31/12. Patient demographic characteristics, health characteristics, and ED visit characteristics were compared to assess differences between those who accept and those who decline testing.

Findings: Of 4510 ED patients offered an HIV test, 3470 accepted for an acceptance rate of 77%. The most common reasons for declining were “no perceived risk” and “tested in the last 3 months.” Those who accepted testing were more likely to be unmarried, less than age 35, Hispanic or African American, Spanish speaking, foreign born, have no primary care provider, report no pain at triage, have a daytime ED visit, and be discharged from the ED compared to admitted. Sex, Employment status, and ED length of stay did not affect whether patients accepted testing.

Principal conclusions: Acceptance of ED-based rapid HIV testing is not universal, and there are both patient and visit characteristics consistently associated with declining testing. This detracts from the goal of using the ED to screen a large number of at-risk patients who do not have access to testing elsewhere.

(C) 2014

Introduction

In 2006, the Centers for Disease Control and Prevention (CDC) modified their recommendations to encourage HIV testing at all points of contact with the health care system, including the emergency department (ED) [1]. A number of factors contributed to the feasibility of these recommendations, including the availability of accurate, rapid, Clinical Laboratory Improvement Amendments (CLIA)-waived, point-of-care HIV tests and increasing acceptance of the ED as an appropriate venue for providing public Health screening

? Sources of Support: This study was funded by a grant from Gilead Sciences, Inc. Gilead reviewed the study design but had no role in the design or conduct of the study or in the collection, management, analysis, or interpretation of the data.

?? Prior presentations: Academy of Emergency Physicians annual meeting, October

2013 in Seattle, Washington.

* Corresponding author. Boston Medical Center, 1 Boston Medical Center Place, Dowling 1 South, Department of Emergency Medicine, Boston, MA 02118. Tel.: +1 617

414 4996; fax: +1 617 414 7759.

E-mail address: [email protected] (E.M. Schechter-Perkins).

[2]. The CDC recommendations included methods to make large-scale HIV testing more streamlined, including opt-out testing, and eliminating pretest counseling.

In response, EDs nationwide began HIV testing programs using a variety of models [3]. Multiple barriers have been overcome as the number of programs increased [4,5]. In 2007 the American College of Emergency Physicians board of directors recommended “HIV testing in the evaluation for acute care conditions in the ED should be available in an expeditious and efficient fashion” [6]. However, studies have consistently found a higher HIV testing refusal rate in the ED than in other venues [7]. At the same time, preliminary evidence indicates that those who decline tests may have significant risk factors for HIV and may be more likely to be HIV positive [8].

Understanding differences between patients who accept and decline HIV testing is important for developing methods to reduce decliner rates among patients who might be at high risk for undiagnosed HIV. The primary objective of this study was to

http://dx.doi.org/10.1016/j.ajem.2014.05.034

0735-6757/(C) 2014

1110 E.M. Schechter-Perkins et al. / American Journal of Emergency Medicine 32 (2014) 1109-1112

determine if there were differences in patient or ED visit character- istics between those who accepted and those who declined HIV testing. Secondary objectives were to examine the rates of acceptance and declining testing, and the reasons for declining testing.

Methods

Study design

This study was a retrospective medical record review of data on all patients offered a rapid HIV test in the ED from 11/1/11 to 10/31/12. The study was approved through expedited review by the Boston University School of Medicine institutional review board, with a waiver of informed consent, and was supported by a grant from Gilead Sciences Inc. Gilead had no role in the design, conduct, or reporting of the study.

Study setting and population

The study population consisted of ED patients 18 years or older, at an Urban academic medical center with 130 000 annual ED visits. The ED is nationally recognized as the primary “safety net” provider for the underserved population in the city; 70% of patient visits are among individuals who rely on government payors for insurance coverage, and 30% of patients have a primary language other than English.

HIV testing program

In September 2009, a collaborative effort between the ED, the Infectious Disease Department, and the Massachusetts Department of Public Health (DPH) resulted in the development of an ED rapid HIV testing program, funded by the DPH. The program employs HIV counseling, testing, and referral (CTR) staff. Screening hours vary for a total of 120 hours per week of coverage, including some evenings and weekends. It utilizes a combination of approaches: (a) a Diagnostic approach in which clinicians refer specific patients for HIV testing because of Clinical concerns, (b) a targeted approach in which CTR staff test patients based on possible HIV risk factors noted in the medical record, and (c) a screening approach in which CTR staff approach ED patients without regard to HIV risk factors. The opt-in test requires written documentation of verbal informed consent and uses a finger-stick point-of-care rapid assay (Oraquick Advance Rapid HIV-1/2 Antibody test). Approximately 300 ED patients are tested each month, and results are documented in the electronic medical record (EMR). The CTR staff maintains a separate database of all patients who are offered HIV testing.

Study protocol

In December 2012, data were compiled for all patients from 11/1/12 to 10/31/12 with an HIV test documented in their EMR, as well as those who declined HIV testing according to data collected by the CTR staff.

Measurements

An electronic query was run on all patients in the database. Variables in the database were grouped into “patient characteristics” and “ED visit characteristics.” The patient characteristics included demographic information (age, sex, ethnicity, primary language, marital status, employment status, birth country, insurance, and homelessness status), primary care provider (pcp), and past medical history. This included an analysis of the International Classification of Diseases, Ninth Revision (ICD9) from the EMR of previous Psychiatric diagnoses, (ICD9 290-312) and further broken down by substance abuse disorders (ICD9 303-305). The ED visit characteristics included pain score, date/time of visit, triage Emergency Severity Index

score, length of ED stay, and disposition. We also examined the reason reported for those who declined testing.

Data analysis

Data were analyzed using SPSS 20 Statistical Software (SPSS Inc., Chicago, IL). Because this was a retrospective chart review of all patients offered HIV testing, no a priori sample size analysis was performed. Descriptive statistics were calculated, and bivariate ?² analysis was used to assess differences between those who accepted and those who declined HIV testing.

Results

During the study period, 3470 of 4510 ED patients offered an HIV test accepted for an acceptance rate of 77%. Twelve of these patients tested positive, for a positivity rate of .3%.

Of those offered HIV testing: the median age was 31 years, 48% were female, and 57% were African American, 23% Hispanic, and 16% white. Eighty-seven percent had public or no health insurance, 46% were employed, and 17% had no usual place of health care.

Overall, acceptance of HIV testing was high. The most common reasons for declining an HIV test were “no perceived risk” (457, 44%), “tested in the last 3 months” (432, 42%), and “would rather test elsewhere” (102, 10%).

Bivariate analysis revealed that both patient characteristics (age, language, marital status, and comorbidities) and ED visit character- istics (pain score and disposition) impacted the likelihood of accepting an HIV test (Tables 1 and 2).

Discussion

Previous studies report acceptance rates of ED Rapid HIV Testing from 40% to 75% when testing is done in the context of an opt-in program as ours is, regardless of whether the program is screening or targeted [9-13]. Acceptance may be higher when testing is offered in an opt-out program [14]; however, legislation mandating either written informed consent or HIV counseling precludes many states from employing this methodology. Our study, which had extremely high acceptance rates of 77% overall, still found that acceptance of ED HIV testing is not universal, and there are both patient and ED visit characteristics consistently associated with declining testing.

In our study, patients who accepted HIV testing were likely to be younger, born outside the United States, nonwhite, and unmarried. This demographic traditionally has poor access to preventative medical care; the finding that having no primary care provider increases likelihood of accepting an ED HIV test is similar to previous studies [10]. This reinforces the appropriateness of the strategy of increasing availability of HIV testing by offering it in the ED. It is important to note that African Americans were 1.6 times as likely to accept compared to whites, because this group is being diagnosed with HIV at a disproportionate rate in the United States [15] and increasing diagnoses of HIV in this group is a target of CDC funding efforts. However, it is disheartening that other potentially high-risk groups that typically lack other sources of health care, including patients who abuse substances and those with psychiatric comorbid- ities, were less likely to accept testing and therefore not being effectively reached by the current model used in our program.

The most common reason reported for declining HIV testing is that the patient felt they had no perceived risk for HIV. This finding has also been reported in other studies of ED HIV testing [11,16; however, in Massachusetts DPH report, 24% of newly-infected individuals were diagnosed with “no identified risk” as the method of exposure, the second largest method of exposure among newly diagnosed

E.M. Schechter-Perkins et al. / American Journal of Emergency Medicine 32 (2014) 1109-1112 1111

Table 1

Characteristics associated with accepting an HIV test

Variable Total	n 4,263	% Accepted	Risk Ratio of Accepting an HIV Test
			Point Estimate	95% CI
Patient Demographic Characteristics
Gender Female Male	2,056 2,207	77% 76%	1 0.98	Reference (0.95-1.01)
Age 18-35 36-64 65+	2,675 1,545 43	80% 71% 44%	1 0.88 0.55
				Reference (0.85-0.91) (0.39-0.77)
Language English Speaker Spanish Speaker Other Language	3,589 391 283	76% 82% 75%	1 1.08 .99
				Reference (1.03-1.14)
				(0.92-1.06)
Birthplace US Born Foreign Born	2,704 1,559	75% 79%	1 1.05
				Reference (1.01-1.08)
Ethnicity White African American Hispanic Other	663 2,433 986 181	68% 77% 82% 74%	1 1.14 1.21 1.10
				Reference (1.08-1.21) (1.14-1.28)
				(.99-1.23)
Marital Status Married Unmarried	465 4,045	71% 77%	1 1.09
				Reference (1.02-1.16)
Insurance Private Uninsured Medicaid Medicare Other Unknown	500 1,171 1,838 296 128 330	75% 82% 76% 72% 79% 67%	1 1.09 1.03 0.83 1.21 0.86
				Reference (1.03-1.16)
				(0.95-1.07) (0.87-1.04) (0.95-1.17) (0.82-.98)
Employment Unemployed Employed	2,345 1,918	76% 77%	1 1.02	Reference (0.98-1.05)
Patient General Health Characteristics
Primary Care Provider Has PCP No PCP	3,570 693	76% 81%	1 1.07
				Reference (1.02-1.11)
Psychiatric comorbidities (by ICD9) None 1 or more disorders	3,131 1,132	78% 73%	1 0.94
				Reference (0.90-0.97)
Substance Abuse (by ICD9) None Psychoactive Substance	3,489 774	78% 71%	1 0.91
				Reference (0.87-0.96)
Charlson Comorbidities (by ICD9) Charlson = 0 Charlson = 2+ Charlson = 1	3,354 452 445	78% 64% 77%	1 0.81 0.99
				Reference (0.76-0.87)
				(0.93-1.04)
ED Visit Characteristics
Pain Score at Triage Pain = 0 Pain >=1	917 3,346	82% 75%	1 0.92
				Reference (0.89-0.95)
ED Length of Stay LOS 0-2 Hours LOS 2-6 Hours LOS 6+ Hours	1,093 2,655 459	77% 78% 76%	1 1.01 0.98	Reference (0.97-1.05) (0.92-1.04)
Number of visits in the last 1 year 1-2 ED visits past yr. 3+ ED visits past yr.	3,408 910	80% 70%	1 0.87
				Reference (0.84-0.92)
Time of Triage Daytime Evening (7pm-11pm)	4,041 221	77% 70%	1 0.91
				Reference (0.83-0.99)
Disposition Discharged Admitted	3,787 476	77% 73%	1 .9	Reference (0.9-1.01)
Triage emergency severity index ESI 5 (No resources) ESI 4 (1 resource) ESI 3 (2 or more resources) ESI 2 (high risk patient) ESI 1 (immediate intervention) Unknown ESI	966 1,622 1,488 129 0 56	73% 78% 80% 70% 0% 0%	1 1.08 1.1 0.96 - -
				Reference (1.03-1.13) (1.05-1.16)
				(0.85-1.08)

Table 2

Reasons for refusing an ED HIV test

Reason n %

Does not want to know test results 41 3.94

No perceived risk 457 43.94

Does not have time 20 1.92

Does not want Blood draw 19 1.83

Tested in last 3 mo 432 41.54

Too tired/feels too ill 57 5.48

Non-English speaking 2 0.19

Known HIV+ 51 4.90

Would rather take test elsewhere 102 9.81

“No reason”	126	12.12
Other	55	5.29
Missing	718

individuals [17]. This suggests that patients are inaccurately inter- preting their risks for HIV infection and may require further education about HIV risks. The increased time required to provide this education and counseling needs to be balanced against the potential benefit of expanding acceptance of testing to a potentially at-risk group that has been shown in our study and others to consistently decline testing.

This is the first study we are aware of that looks at the association between testing acceptance and various ED visit characteristics. Certain visit characteristics were associated with acceptance of testing, including reporting no pain at triage; an ESI triage score of 3 or 4 rather than 1 or 5; and being discharged from the ED rather than admitted. These characteristics are generally associated with being less sick at the time of test offer. The exception is a decreased likelihood of accepting among those with ESI triage of 5; we suspect that this group is likely to be in the ED for a short time and thus maybe unwilling to wait for an HIV test.

We found that daytime ED visits were more likely to result in test acceptance than evening visits. This is in contrast to a previous study, which found that patients with evening ED visits were more likely to accept testing [11]. It is difficult to reconcile these findings because both studies offered testing in the day and evening. More research is needed to determine which periods merit investing limited resources to maximize testing acceptance.

This study had several limitations that need to be kept in mind when interpreting the findings. First, the study was a retrospective review, and data collected were limited to medical record documentation; it is possible that additional patients declined testing but were not recorded by CTR staff. Second, there is no defined protocol regarding which patients are offered HIV tests; counselors may target patients they believe likely to accept testing in addition to those at high risk for testing positive, thus artificially inflating the acceptance rate. CTR staff are primarily charged with testing a defined number of patients each month, and this charge may be in direct conflict with the objective of offering testing to all patients. Third, the study was conducted at a single academic institution, and results might not be generalizable. Finally, this study presupposes that undiagnosed HIV is a possibility in those who refuse HIV testing, although in reality, we do not know the risk of HIV among those patients.

Our program achieved a high acceptance rate for ED HIV testing; however, even with this rate, there are consistent high-risk groups of patients not reached with the current model. The ability of the ED to screen and test a large number of at-risk patients who do not have access to testing elsewhere is thus diminished. As the national discussion about HIV testing moves toward finding sustainable ways to diagnose HIV early in an era of limited resources, strengthening ED HIV testing programs will require programmatic choices about whether to focus on targeting patients that are likely to accept testing, or to diverge from the CDC recommendations and provide counseling to patients that decline to encourage them to accept the ED HIV test.

Acknowledgments

Monica Malowney, Gina Lee, and Mahima Mangla are acknowl- edged for their help in data collection.

References

1. Bernard M, Branson MH, Hunter Handsfield M, Margaret A, Lampe M, Janasen R, Taylor A, Lyss S, et al. Revised recommendations for HIV testing of adults, adolescents, and pregnant women in health-care settings. MMWR 2006;55:1-17.
2. Irvin CB. Public health Preventive services, surveillance, and screening the emergency department’s potential. Acad Emerg Med 2000;7(12):1421-3.
3. Haukoos JS, White DAE, Lyons MS, et al. Operational methods of HIV testing in emergency departments: a systematic review. Ann Emerg Med 2011;58:S96-S103.
4. Mumma BE, Suffoletto BP. Less encouraging lessons from the front lines: barriers to implementation of an emergency department-based HIV screening program. Ann Emerg Med 2011;58:S44-8.
5. Arbelaez C, Wright EA, Losina E, et al. Emergency provider attitudes and barriers to universal HIV testing in the emergency department. J Emerg Med 2012;42(1):7-14.
6. HIV testing and screening in the emergency department. Ann Emerg Med 2007;50:209.

   1112 E.M. Schechter-Perkins et al. / American Journal of Emergency Medicine 32 (2014) 1109-1112

   Rapid HIV. Testing in emergency departments-three U.S. sites, January 2005- March 2006. MMWR Morb Mortal Wkly Rep 2007;56(24):597-601.

7. Czarnogorski M, Brown J, Lee V, et al. The prevalence of undiagnosed HIV infection in those who decline HIV screening in an urban emergency department. AIDS Res Treat 2011 [[Internet], [cited 2013 Sep 4];2011. Available from: http://www. hindawi.com/journals/art/2011/879065/abs/].
8. Merchant RC, Seage GR, Mayer KH, Clark MA, DeGruttola VG, Becker BM. Emergency department patient acceptance of opt-in, universal, rapid HIV screening. Public Health Rep 1974-2008;123:27-40.
9. Bamford L, Ellenberg JH, Hines J, Vinnard C, Jasani A, Gross R. Factors associated with a willingness to accept rapid HIV testing in an urban emergency department. AIDS Behav 2013:1-4.
10. Pisculli ML, Reichmann WM, Losina E, et al. Factors associated with refusal of rapid HIV testing in an emergency department. AIDS Behav 2011;15(4):734-42.
11. Lyons MS, Lindsell CJ, Ruffner AH, et al. Randomized comparison of universal and targeted HIV screening in the emergency department. J Acquir Immune Defic Syndr 2013;64.3(2013):315-23.
12. Trotter AB, Bhayani N, Florsheim R, Novak RM. Implementing universal oral HIV screening in an urban emergency department-do demographic characteristics impact acceptance of testing? Clin Infect Dis 2010;50(2):283.
13. Freeman AE, Sattin RW, Miller KM, Dias JK, Wilde JA. Acceptance of rapid HIV screening in a southeastern emergency department. Acad Emerg Med 2009;16 (11):1156-64.
14. HIV Surveillance Report, vol. 23; 2011.
15. Brown J, Kuo I, Bellows J, et al. Patient perceptions and acceptance of routine emergency department HIV testing. Public Health Rep 2008;123(Suppl 3):21-6.
16. Massachusetts State HIV/AIDS Plan. [Internet] Available from: http://www.mass. gov/eohhs/docs/dph/aids/mass-hiv-aids-plan.pdf; 2013.